Prognostic Utility of Glucose-to-Potassium, Lactate-to-Albumin, and Uric Acid-to-Albumin Ratios for Adverse Outcomes in Acute Upper Gastrointestinal Bleeding: A Retrospective Cohort Study

Background Acute upper gastrointestinal bleeding (UGIB) is a life-threatening emergency with an annual incidence of 50–150 per 100,000 and ~ 10% in-hospital mortality. While scores like AIMS65, Glasgow-Blatchford (GBS), and Rockall aid risk stratification, their limitations in resource-limited settings highlight the need for simple, laboratory-derived biomarkers. This study evaluates admission glucose-to-potassium (GPR), lactate-to-albumin (LAR), and uric acid-to-albumin (UAR) ratios for predicting mortality, ICU admission, and intubation in endoscoped emergency department (ED) patients with UGIB. Methods In this retrospective cohort study at a tertiary hospital (July 2024–July 2025), 472 adults (≥ 18 years) with confirmed UGIB (hematemesis/melena/hematochezia) undergoing endoscopy were included. Baseline demographics, vitals, labs, and scores (AIMS65, GBS, Rockall) were recorded. Ratios were calculated from admission labs. Outcomes were in-hospital mortality, ICU admission, and intubation. Non-normal data were analyzed via Mann-Whitney U and Spearman correlations; predictive performance via ROC curves and DeLong's test (SPSS v20.0; p < 0.05). Results Median age was 71 years (62.3% male); comorbidities included hypertension (47.5%) and diabetes (26.7%). Mortality occurred in 11.2%, ICU admission in 78.0%, and intubation in 11.4%. Non-survivors had higher LAR (0.130 vs. 0.049, p < 0.001) and UAR (0.203 vs. 0.157, p < 0.001). LAR correlated positively with scores (r = 0.414–0.424, p < 0.001). For mortality, LAR AUC was 0.79 (cutoff 0.05; sensitivity 79.2%, specificity 64.2%), matching AIMS65 (0.79) and outperforming GBS (0.65) and Rockall (0.75). Similar for intubation (LAR AUC 0.79). GPR/UAR showed modest AUCs (0.59–0.66). Conclusions LAR provides AIMS65-equivalent prognostic accuracy for mortality and intubation in UGIB, offering a cost-effective adjunct for ED triage. Integrating these ratios may optimize resource allocation and outcomes in high-risk patients.